Art Direction · UI/UX · Design System · E-commerce · iOS App
Helen Marlen — luxury that finally flows online.
Two chapters, years apart: the vision, and the return. Helen Marlen perfected luxury offline — showrooms, personal stylists, try-on at home — but online felt cold. First I art-directed an editorial-commerce vision; when it was scaled down to a lean store, I came back years later to rebuild the design system and make buying effortless, across web and a native iOS app.
Background
A 30-year offline ritual that online couldn't carry.
Helen Marlen is Ukraine's premier luxury retailer — 300+ labels across Woman, Man and Kids, an official retailer since 1994. Offline it is showrooms, personal stylists and try-on at home. Online it was cold product grids, and an iOS app with no system of its own. The brief: the screen should feel like the salon — the service and the visual language.
Challenge
Three shoppers, three fears — and one they all share.
- Losing the personal touch. The loyal salon client didn't want to trade her stylist for a grid.
- Buying a fake. The new affluent shopper needed proof the store is official before paying.
- Overspending. The young shopper couldn't see why one piece beats ten cheap ones.
Underneath all three sat the same question: "what if it doesn't fit?"
Discovery
Discovery before design.
With the in-house UX team we ran a customer survey, client interviews and competitor research, then built personas, framed the hypothesis and moved through wireframes and sign-off before any UI. I took part in the research and led the answer in the interface. Alongside it ran a UI discovery: Vogue-grade editorial, a serif-and-grotesk pairing, black and white. Fit anxiety came up again and again — so fit became a design problem, not a returns policy.
| Who | Fears | Pains | Buys when |
|---|---|---|---|
| Iryna, 42 · loyal salon client | Losing the personal treatment; fit | Cold grids, no one who knows her taste, returns | Her manager continues online, fit is de-risked, the site feels like the boutique |
| Maksym, 38 · new affluent shopper | "Are these original?" — and fit | Can't inspect before paying; unfamiliar stores | The retailer proves it's official, delivery is clear, returns are easy |
| Andrii, 26 · young aspiring shopper | Spending big on one piece | Luxury feels gated; fast fashion dies in a season | He sees cost-per-wear and can earn his way to a first real piece |
Thinking · Chapter 1 — the vision
An editorial feed that learns your taste — not a form.
As Art Director at Vintage I led the UI, the visual language and the design system. The store was rebuilt in the brand's own register — editorial serif over a quiet grotesk UI, campaign photography, room to breathe — and reframed as a feed rather than a catalogue. Preference capture was gamified: skip it at sign-up, meet it as playful prompts while you browse, and the feed personalises. The component library used variables and tokens across 320, 768 and 1280 — the base the iOS app later grew from.
| Fear | The design answer | Status |
|---|---|---|
| "Will it still feel personal — and fit?" | Concierge stylist delivery and try-on at home; an account that remembers sizes and taste; the boutique's visual language | Live today |
| "Is it authentic — and will it fit?" | Official-retailer proof, one price online and offline, measurements and size guide, try-and-pay pickup, 14-day returns | Live today |
| "Is one piece worth ten cheap ones?" | An editorial feed that builds taste; gamified rewards toward a first piece | Planned — covered today only by loyalty, Trade-In and Outlet |
Reality: the scale-down
The feed and the game layer were heavier to build than the brand could resource. They were paused, and the store shipped as a lean, conventional MVP — and with them went the push for the young shopper, which needed communication the brand wasn't ready for. This case shows the full vision next to what shipped.
Outcome · Chapter 2 — the return
Years later they came back — to a mess. I rebuilt the order.
The brief was registration and checkout. The Figma files had drifted into chaos, so before redesigning anything I gathered every screen and rebuilt the design system from scratch — one clean, tokenized source. Then every step between "I want this" and "it's ordered" got shorter.
| Flow | What changed |
|---|---|
| 1 · Sign-in | A 3-step, ~7-field form became a 3-tap phone-code login, plus one-tap social. |
| 2 · Buying | Guest checkout with no account, and Buy it now on the product page that skips the cart. |
| 3 · Checkout | One page: cart, delivery, payment, with the order summary always in view. |
| 4 · Delivery | Stylist concierge, showroom try-on, Nova Poshta or standard — chosen inside checkout. |
Buy without an account
Buy it now opens one short choice: order without registration, sign up or log in. The account becomes an option, not a gate.
One checkout page
Cart, delivery and payment in one place. The try-on and stylist options sit in the delivery step — the brand's service, surfaced where the decision is made.
The app: log in with a code
I inherited an inconsistent iOS app with no system. Sign-in moved to phone number and a confirmation code, with every state drawn: resend timer, wrong code, success — and the email path with its own errors.
Every screen, on one system
~14 account screens and states — home, personal data, orders, wishlist, address book with an empty state, change phone, callback, contact, search with live highlighting — plus a product page aligned to web. Web and app now draw from the same tokens.
What shipped
Results
The service shipped — the feed waits for its moment.
For the loyal and the affluent client the fears are closed on the live store: concierge, try-on, proof of authenticity, easy returns, and a checkout that doesn't ask for an account. What didn't ship is the part built for the young shopper — the editorial feed and gamified taste-building. The live store runs lean as an MVP.
Effortless entry
A 3-step, ~7-field form became a 3-tap phone-code login, one-tap social or guest checkout.
What changed→What changedFrom form to three taps- Before: name, surname, email, phone, password, birthday, gender — plus sizes and favourite brands
- After: phone number → code → in
- One-tap social sign-in
- Guest checkout, no account needed
- Buy it now skips the cart
One system, two platforms
A single tokenized design system for web and native iOS, replacing drifting files.
What it covers→What it coversOne source of truth- Components with variables and tokens
- Web at 320, 768 and 1280
- The iOS app rebuilt on the same tokens
- Every screen gathered before redesign
Rebuilt at depth
~14 account screens and states in the app; three responsive breakpoints on web.
What was rebuilt→What was rebuilt~14 screens and states- Account home, personal data
- Orders and a single order
- Wishlist; address book with an empty state
- Change phone with confirmation
- Callback, contact us
- Search with live query highlighting
The differentiator, surfaced
Try-on and stylist delivery pulled into checkout — aimed at higher order value and fewer size returns.
Where it lives→Where it livesThe salon, inside checkout- Stylist concierge delivery
- Try-on at a showroom
- Nova Poshta
- Standard delivery
Next chapter
- Test the checkout. Usability-test the phone-code and guest paths to find the friction left.
- Measure the fit tools. Track whether richer size guidance lowers size-driven returns.
- Scale the system. Extend the shared library as new features ship.
- Revisit the feed. When the brand can resource the communication the young shopper needs.
Every number here is counted from the design files. No sales data was shared, so no conversion figures are shown.
Selected screens
One editorial system — web and iOS speaking the same language.
